The most suitable method to measure small value resistances is the four wire method. The circuit is simple: you force constant current I through the wire and measure the voltage across the wire by a sensitive voltmeter.

An important point is to make the voltmeter contact inside the current source contact to avoid the voltage drop on the contacts of the current source to the wire under test.

If you have a wire with 4 milli ohm, then if you have millimeter then you have to pass only one ampere in the wire. If this current is too large for the wire to withstand you have to use a more sensitive voltmeter. If you have a microvoltmeter then you need only to pass one mA current in the wire.

In order to set the current you need to use a power supply voltage say of one volt and connect in series with it proper resistance. For one ampere, you need to connect only one ohm in series with it. If you cam measure at only one mA , you have to connect one kiloohm resistance in series with voltage power supply.

Best wishes

There are available on the market microohm meters that can measure such small values. I'm using MI 3252 MICROOHM 100A from Metrel that can measure in range of 1 nΩ ... 20 Ω. The only problem is that with smaller values to measure you need higher current values. That will lead to change in the resistance due to temperature rise. You would need to find the right balance between current value, time of measurement and its resolution.

There is the Wheatstone bridge.

Regards,

Joachim

I am sorry, corrected the value to 4.9 ohm

The variation is in the 0.1-10 milliohm range.

The beth method is using a 4 wire system or weatsone bridge. The weatstone has the problem that it is needed a resistor in similar values to the measured one, but adjust resistors has lots of ppm of error.

The wire is made of Nichrome, so there is low variation due temperature.

It can not be used tin soldering as long as tin can introduce 20 milliohm error.

In order to not vary very much the temperature the current should be injected in pulse mode and fast sample/hold

I have seen MI3252 but is not pulsed one.

I have seen also 444-2 but has only 4 1/2 digit.

The fastest method for me is using a circuit for pulse+sample/hold then measure with a high accuracy multimeter

I think the Wheatstone Bidge is the most suitable method. Also you could just use a really long wire. (1000-10000m).

I used the 4 wire method many times to measure small and very small value resistances. The main advantage is that you do not need to bond the any of the 4 access wires. You can use just a pressure contact by a crocodile. The two outer wires are connected to the current source and the two inner wires are connected to the voltmeter. In this way the voltage drop on the outer contacts to the current source will be excluded. Since the voltmeter has high resistance, the contact resistance of its connecting wires will not drop any appreciable voltage.

Th 4-wire electrode system is a very suitable method to measure small and very small resistances.

The bridge which is suitable to measure small resistance with sufficient accuracy is the Kelvin bridge.

Best wishes

After reading about advantages of the 4 wires probe and Wheatstone, I should use a mix of 4 wire and an adapted voltage reference (similar to a wheastone but measuring voltage directly instead of current):

I can inject a DC current, as said Victor Lorenzo, and measure again an accurate voltage reference at same voltage than the initial voltage measured at the 4 wires probe to measure the voltage difference, as proposed attached diagram.

I attachedalso a high current source "almost stable" that I have simulated

Wheatstone Bridge is the best way to measure it.

How can we used the Wheatstone bridge? can you show on a figure?